No Card Network Rule Assigns AI Agent Purchase Liability to Platforms

Merchants currently bear default liability when AI agents make mistaken or unauthorized purchases, with no Visa or Mastercard rule assigning responsibility to an AI platform. A coalition of banks published a report warning about this gap, noting that AI shopping agents entering card details directly expose merchants to chargeback liability under existing rules. Card networks have built technical standards for carrying an intent mandate from an agent to a merchant, but no specific AI-agent liability rules have been published by card brands.

Merchants and platforms are responding differently. Target holds users 100% liable for anything an agent does. Expedia declared on 24 September that it stays merchant of record for hotel bookings made via its Muse agent, keeping dispute ownership with the merchant. Amazon is blocking AI agent access, including Muse. Satya Nadella argued that insurance and market pricing will ultimately resolve liability questions, saying technical protocols for agents are good but that market pricing is the ultimate arbiter. The Merchant Advisory Group noted that some AI agents may absorb liability for errors or hallucinations as a premium commercial feature, though it also noted this flow may become prohibitively expensive for agents to manage. Authentication systems currently struggle to distinguish AI agents from malicious bots, complicating buyer verification.

On the consumer side, Traxtech reported that 73% of shoppers use AI in their purchasing journeys and 70% are comfortable having AI agents make purchases on their behalf, with payment security the top concern at 32%.